Differential scanning calorimetry of aluminium EN AB-42000 alloy rheocasting semi-solid in different stage heating rates

The EN AB-42000 alloy has a melting temperature of 680 °C. The phase diagram depicts the influence of AlSi7 weight per cent and phase transition temperatures [8]. The alloy was melted at 700 °C in an electric resistance furnace [10]. To allow for melting temperature, the alloy was kept in this furnace for minutes. The crucible was removed from the furnace, and the molten metal was placed in a stainless steel cylinder cup, which was permitted to cool at room temperature. The procedure outlined above was repeated for the same metal during the rheocasting process. Once the crucible was removed from the furnace, the melted metal was internally cooled using a ‘K type’ thermocouple to regulate the temperature. Subsequently, the molten metals were vigorously agitated for a few seconds at 640 °C using steel attached to a spinning electric hand drill. This agitation was performed until the melt completely dissolved, aiming to produce globular shapes and disrupt the dendritic structure. To prevent dendritic growth, the alloy should be cooled to room temperature with water as soon as the dendritic microstructure is replaced with a globular microstructure [30]. DSC has been used in several studies to investigate solid-state phenomena in aluminium alloys, such as precipitation and dissolution of precipitates, as well as precipitation kinetics during heat treatments [24]. The specimen under inquiry and a reference specimen is heated and cooled at the same time in a DSC experiment [34]. Within the examined temperature range, the reference specimen must not display any phase change [35]. During such a heating cycle, the difference in energy consumption or output per time and mass between the specimen and the reference is measured. As a function of temperature, this so-called heat flow is displayed. An exothermic process, such as solidification or precipitation, is indicated by a positive peak. An endothermic process, such as melting or dissolution, is represented as a negative peak [36]. The semi-solid rheocasting offers several advantages over conventional casting techniques, making it especially suitable for aluminium alloys like EN AB-42000. These benefits include improved flow ability, reduced defects (like porosity and hot tearing), enhanced mechanical properties, and lower energy consumption. Furthermore, rheocasting is cost-efficient, producing near-net-shape parts requiring less machining. These advantages enhance the performance and viability of components made from EN AB-42000, a widely used alloy in demanding sectors like automotive, aerospace, and electronics industries [37]. Further details can be found in the authors previous studies [33].

The rheocasting can offer additional and valuable insights into the solidus temperature of an alloy, which can complement or enhance the information obtained from traditional techniques like thermodynamic calculations or differential scanning calorimetry (DSC) analysis. The advantages of rheocasting include its ability to reflect real-world conditions, its use of larger, more representative samples, the possibility of visual observation, and the opportunity for post-process microstructure analysis. However, rheocasting is not a replacement for other methods but should be used in conjunction with them to obtain a comprehensive understanding of an alloy's behaviour.

Smaller sample sizes are preferred for DSC measurements to ensure even heating throughout the sample and a quicker response to changes in temperature. Larger, non-homogeneous samples may heat unevenly, and their thermal inertia might slow down responses to temperature changes, leading to inaccurate results.

Similarly, slower heating rates are advantageous as they allow sufficient time for phase transformations to occur at their "proper" temperatures. Faster heating rates may cause these transformations to appear at higher temperatures and can also reduce the resolution of closely spaced thermal events.

The use of smaller sample weights and slower heating rates can lead to sharper, better-defined peaks in the DSC curve and a more accurate determination of phase transition temperatures. This improves the understanding of the alloy's thermal behaviour and makes it easier to interpret different thermal events. As a result, this approach can provide valuable insights for optimizing processes like rheocasting. Table 2 shows the number of DSC tests, including samples, measurements of alloy tape, sample dimension, and crucible type utilized in the test.

To gain more accurate insights, researchers can mitigate these challenges by conducting multiple experimental runs, utilizing complementary characterization techniques, and applying theoretical calculations.

The measurement sample was two-stage heating rates to 550 °C and fast then slow heating, in the second stage fast cooling sample. The thermal cycle for the first measurement started heating to 500 °C at 100 °C/min, heating to 550 °C at 50 °C/min, holding at 550 °C for 1 min, and cooling to RT at 100 °C/min.

In Fig. 1a, the analysis measurement of the sample is displayed. We can observe a small exo peak around 350 °C and a sudden drop in heat flow at 500 °C when there’s a change in the heat rate. Control stabilises as the range increases by about 20 °C, and rapid cooling ensues. However, control is lost at 350 °C, making the cooling rate of 100 °C/min unsuitable for results and sample inspection. Notably, no melting occurs in this instance.

Conversely, Fig. 1b provides an analysis measurement where the small exo peak at around 350 °C is absent. The endo-reaction begins at approximately 570 °C, suggesting potential melting. During cooling, a prominent exo peak indicates solidification. A cooling rate of 50 °C/min is applicable for this measurement. Upon sample inspection, there was a noticeable color change, indicating that some melting might have taken place. The sample did not stick to the holder. The second measurement of the first sample was fast heating to 600 °C. The thermal cycle for the second measurement started heating to 600 °C at 50 °C/min, heating to 600 °C at 50 °C/min, holding at 600 °C for 1 min, and cooling to RT at 50 °C/min.

In a DSC analysis, when the heating or cooling rates are high (for instance, 20 °C/min), the rapid energy changes that occur during phase transformations might exceed the measurement limit of the DSC instrument. This limitation can lead to inaccurate enthalpy measurements, difficulty in distinguishing thermal events, and potential misinterpretation of the alloy's thermal behaviour. Several steps can be taken to mitigate these limitations: reducing the heating or cooling rate can result in more accurate measurements; using smaller samples can decrease the total energy change, reducing the likelihood of exceeding the measurement limit; using a high-capacity DSC instrument with higher measurement limits can handle larger heat flows.

The measurement of the sample was fast heating to 650 °C and cooling was not important. The thermal cycle for the third measurement started heating to 650 °C at 50 °C/min, holding at 650 °C for 1 min, and cooling to RT at 100 °C/min.

Figure 2a shows an analysis third measurement of the first sample was a small exo peak at around 350 °C, endo-reaction starts at about 570 °C, with a small peak, then a large one. Melting might not be completed, a cooling rate too high, not measurable, and sample inspection; colour changed, some melting might occur. The sample did not stick to the holder.

Figure 2b demonstrates an analysis measurement of the sample where a small exo peak at around 350 °C, endo reaction onset at around 570 °C, with a small peak, then a large one, large, elongated peak-end at about 670 °C, complete melting captured, a cooling rate too high, not measurable and sample inspection; colour changed, some melting might occur. The piece did not stick to the holder.

The fourth measurement of the first sample was fast heating to 700 °C. The thermal cycle for the fourth measurement started heating to 700 °C at 50 °C/min, holding at 700 °C for 1 min, and cooling to RT at 100 °C/min.

The measurement of the sample aimed to control heating to 700 °C. The thermal cycle for the fifth measurement started heating to 700 °C at 20 °C/min, holding at 700 °C for 1 min, and cooling to RT at 100 °C/min.

Figure 3a and b illustrates the analysis measurements of the sample. A small Exo peak appears at around 350 °C, followed by an endo reaction onset at approximately 570 °C. This is initially marked by a small peak, which then escalates to a more significant peak ending around 650 °C, indicating complete melting. Due to the excessively high cooling rate, the measurement couldn't be taken. Upon inspecting the sample, a change in color was observed, suggesting that some melting might have occurred. The sample did not stick to the holder.

The measurement of the sample aimed to control heating to 700 °C, and control cooling; repeated cycle. The thermal cycle for the sixth measurement started heating to 700 C at 20 °C/min, holding at 700 °C for 1 min, cooling to 300 at 40 °C/min, holding at 300 °C for 1 min, heating to 700 °C at 20 °C/min, hold at 700 °C for 1 min and cooling to RT at 40 °C/min.

Figure 4a Figure 4a illustrates the analysis measurement of the sample. The results indicate that complete melting occurred, and the cooling rate was deemed satisfactory.

Measured elongated exo peak during solidification onset just below 620 °C; a small endo peak at the middle, sample inspection, the colour changed, some melting might occur, and the sample did not stick to holder.

Figure 4b presents the analysis measurement of the sample. At the observed rate, we can identify 2 distinct peaks. The onset of the endo reaction starts around 570 °C, marked by a minor peak. This is followed by a more prominent peak indicating initial melting. A subsequent peak, culminating at about 650 °C, signifies complete melting. The cooling rate was found to be satisfactory.

Measured elongated exo peak during solidification onset at 630 °C; a second peak ended at 565 °C, for heating. The liquid fraction is calculated by integrating the peaks. For cooling, the solid fraction is calculated by integrating the peaks, sample inspection: colour changed, some melting might occur, and the sample did not stick to the holder. The aim of the measurement of the sample was slow heating to 700 °C and slow cooling; repeated cycle. The thermal cycle for the seventh measurement started heating to 700 °C at 5 °C/min, holding at 700 °C for 1 min, cooling to 300 at 5 °C/min, and cooling to RT.

The aim of the eighth measurement of the third sample was cycling temperatures between 500 and 700 °C with different rates (0.5, 1, 2, 5, 10 and 20 °C/min), recording heating and cooling, calculating solid fractions

Figure 5 demonstrates an analysis measurement of the sample was heating and cooling, calculating solid fractions. In our analysis, we observed that the onset of melting occurs around 590 °C and has limited dependence on the heating rate. For each heating rate, heat flow curves consistently follow the same pattern. However, the conclusion of the melting process is strongly tied to the heating rate, with peak heights rising correspondingly. Notably, at heating rates of 10 and 20 °C/min, the heat flow surpasses the measurable range. Additionally, a minor peak usually seen at 580 °C is absent at these rates. The cooling analysis cooling peaks shift to the left, the onset of solidification is dependent on the cooling rate, and the ending of solidification is highly dependent on the cooling rate. Upon inspecting the specimen, a darkened surface with indications of melting was observed. The diameter appears to have increased, yet the specimen can still be removed from the holder.

Figures 6, 7, 8 demonstrate an analysis measurement of the sample was heated rates at 10 and 20 °C/min; peaks were smeared. High values exceeding measurement limits, at lower rates, 2 separated peaks for the eutectic reaction and complete melting temperatures, even at low rates, heating, and cooling peaks were shifted; repeated cycles show good measurement reliability.

The first heating/cooling and first reheat cycles at 0.5 °C/min show some minor deviation, indicating a difference between the initial state and the reheated state of the sample, and there is a breakpoint in the solid fraction curves at about fs = 60%. There is a significant difference between the S parameter (S = df/dT) S = 0.6/10 (1/°C) in the range of fs = 0–60%; and S = 0.4/40 (1/C) at fs = 60–100%.

Table 3 illustrates an analysis measurement of the sample where characteristic temperatures are the onset of the transformation, a break-point, and the ending. All temperatures show an increasing trend with the rising heating rates; partially from the physical fact of material behaviour, and partially from the thermal lag of the measurement method and the increase of the transformation energy (e.g. total latent heat) shows also an increase with the heating rate; a deviation was seen at 10 and 20 °C/min due to the exceeded measurement limit.

The aim of the measurement of the sample was cycling between 500 and 700 °C with different rates (0.5, 1, 2, 5, 10, 20 °C/min), recording heating and cooling, and calculating solid and liquid fractions. (Fig. 9) demonstrates a ninth analysis measurement of the third sample was heating and cooling, calculating solid fractions. In analysis, two peaks can be identified, the low-temperature peak is for the melting/solidification of the eutectic, and the high-temperature peak is for the melting/solidification of the alpha phase.

The two peaks may overlap depending on the heating and cooling rate. On heating rate, the onset of melting is less dependent, at about 590 °C; at each heating rate, heat flow curves follow the same envelope, ending of melting is highly dependent on the heating rate, Peak heights increase with the heating rate, at 10 °C/min and 20 °C/min heat flow exceeds the measurable range, tiny peak at 580 °C is missing. On cooling rate, cooling peaks shift to the left, the onset of solidification is dependent on the cooling rate, and ending of solidification is highly dependent on the cooling rate. Inspection of the specimen reveals darkened surface with signs of melting; the diameter might have been increased; nevertheless, the specimen can be taken out of the holder.

Figures 10, 11, 12 demonstrate an analysis measurement of the sample was heated rates at 10 and 20 °C/min, peaks are smeared, and high values exceeding measurement limits, at lower rates, two separated peaks for the eutectic reaction and complete melting temperatures.

The observed enthalpies during the differential scanning calorimetry (DSC) analysis of the aluminium alloy EN AB-42000 are within the range of 360–430 mJ/g. There are observable trends, but drawing firm conclusions from them is challenging due to the complexities of the system. Two key trends were observed:

Even at low rates, heating, and cooling peaks are shifted, and repeated cycles show good measurement reliability. The first heating/cooling and first reheat cycles at 0.5 °C/min show some minor deviation, indicating a difference between the initial state and the reheated state of the sample. There is a breakpoint in the liquid fraction curves at about fl = 60%. There is a significant difference between the S parameter (S = df/dT) S = 0.6/10 (1/°C) in the range of fs = 0–60%; and S = 0.4/40 (1/°C) at fs = 60–100%. The breakpoint at cooling is at fs = 50%.

Tables 4 and 5 illustrate an analysis measurement of the sample where characteristic temperatures are the onset of the transformation, a breakpoint, and the ending of the transformation.

Extracted data, at heating, all temperatures show an increasing trend with the rising heating rates: partially from the physical fact of material behaviour and partially from the thermal lag of the measurement method. The increase of the transformation energy (e.g. total latent heat) also shows an increase in the heating rate; a deviation is seen at 10 and 20 °C/min due to the exceeded measurement limit. Extracted data, at cooling, all temperatures show a decreasing trend with the increasing heating rates, partially from the physical fact of material behaviour and partially from the thermal lag of the measurement method. There is a decrease in the transformation energy with increasing cooling rates; the thermal capacity of the sample holder may cause a reduction in the apparent heat measured; a deviation is seen at 20 °C/min due to the exceeded measurement limit.

The aim of the measurement of the sample was cycling between 500 and 700 °C with different rates (0.5, 1, 2, 5, 10, 20 °C/min), recording heating and cooling, calculating solid and liquid fractions, the sample Ambient N, 20 ml/min. (Fig. 13) demonstrates a tenth analysis measurement of the fourth sample was heating and cooling, calculating solid fractions with multi-cycle with reduced mass. The sample weight is about 2/3 of the previous measurements (~ 80 mg vs. ~ 120 mg). Compared to sample 3 results, peak heat flows decreased, but the measurement limit still exceeded 20 °C/min. The total temperature window in the plotted range is 540–660 °C. The repeatability of the measurement is less accurate for the lower rates, especially for the solidification of the alpha phase. Inspection of the samples shows a large amount of oxidization.

Figures 14, 15, 16 demonstrate an analysis measurement of the sample was still 10 and 20 °C/min, peaks are smeared, and high values exceed measurement limits.

The two reactions are distinguishable; even at low rates, heating, and cooling peaks are shifted.

The first heating/cooling and first reheat cycles at 0.5 °C/min show some minor deviation, indicating a difference between the initial state and the reheated state of the sample.

Tables 6 and 7 illustrate an analysis measurement of the sample where characteristic temperatures are the onset of the transformation, a breakpoint, and the ending of the transformation. At heating, at 0.5 °C/min curves slightly deviated from the others, as the sample is in the initial state, second heating at 0.5 °C/min shifts the heating curve to the right, and reheated cases show more similarities in their trends.

There is a slight increasing trend in the onset temperature; enthalpy is distinctly larger for the first heating, the similar for 1 and 2 °C/min, maybe slightly decreasing; data at 10 and 20 are not suitable for comparison due to the exceeded measurement limit in heat flow during the eutectic reactions. At cooling, heat flow deviates at 0.5, 1, and 2 °C/min, but transformation temperatures are very similar, while enthalpy shows an increasing trend. All temperatures show a decreasing trend with the rising heating rates; a deviation is seen at 20 C/min due to the exceeded measurement limit; enthalpy is increasing with the increasing cooling rates.

The aim of the measurement of the sample was cycling between 500 and 700 °C with different rates (0.5, 1, 2, 5, 10, 20 °C/min), recording heating and cooling, calculating solid and liquid fractions, and the sample Ambient N, 20 ml/min. (Fig. 17) demonstrates a tenth analysis measurement of the fifth sample weight is further reduced to ~ 40 mg (vs. ~ 80 mg and ~ 120 mg). Compared to sample 4 results, peak heat flows further decrease; the measurement limit covers the largest heating/cooling rates (20 °C/min) as well. The total temperature window in the plotted range is 540–660 °C; the repeatability of the thermal cycles is quite accurate. Inspection of the samples shows a large amount of oxidization; it shows that the bottom is melted; the top of the specimen keeps the original structure, and seemingly the top shell is not melted.

Figures 18, 19, 20 demonstrate an analysis measurement of the sample was the two reactions are well distinguishable all over the heating–cooling rates.

The first heating/cooling and first reheat cycles at 0.5 °C/min show some minor deviation, indicating a difference between the initial state and the reheated state of the sample. The next reheating shows good repeatability.

Figure 21 illustrates an analysis measurement of the sample where characteristic temperatures are the onset of the transformation, a breakpoint, and the ending of the transformation. At heating, at 0.5 °C/min, curves slightly deviated from the others, as the sample is in the initial state. Second heating at 0.5 °C/min shifts the heating curve to the right; reheated cases show more similarities in their trends.

There is a slight increasing trend in the onset temperature. Enthalpy is distinctly larger for the first heating, similar for 1 and 2 °C/min, maybe slightly decreasing; data at 10 and 20 are not suitable for comparison due to the exceeded measurement limit in heat flow during the eutectic reactions. At cooling, heat flow deviates at 0.5, 1, and 2 °C/min, but transformation temperatures are very similar, while enthalpy shows an increasing trend. All temperatures show a decreasing trend with the rising heating rates; a deviation is seen at 20 °C/min due to the exceeded measurement limit. Enthalpy is increasing with the increasing cooling rates.

Figure 21 demonstrates an analysis measurement of the sample was latent heat trends vs. the heating/cooling rate; the data extraction from the last measurement set (Al5, m = 40.1 mg) is the most accurate, and it shows reliable trends. Enthalpy for first heating and cooling deviates from the subsequent thermal cycles. In the subsequent thermal cycles, reproducibility is good. Enthalpies in heating with increasing rates are decreasing, enthalpies in cooling with rising rates are growing, and enthalpies are asymptotic towards a value a bit higher than 400 J/g (~ latent heat for pure Al = 396 J/g). Latent heat for AlSi7 in the literature is at 480–500 J/g.

Figures 22, 23 demonstrate an eleventh measurement of the sample each peak moves with a different speed concerning the heating/cooling rates. At 3 heating and cooling, peaks would be located at identical temperatures at the intersection of the corresponding lines in the diagram.

The intersection of eutectic peaks in (Fig. 23) (p1 heating, and p4 cooling) would be at a heating/cooling rate of 0.17 °C/min, and the intersection of alpha peaks (p2 heating, and p3 cooling) would be at a heating/cooling rate of 0.39 °C/min.